The Central Florida Regional Transportation Authority, commonly known as Lynx, started providing service on an improved downtown circulator, Lymmo, on August 4, 1997. The service offers the following features:
- exclusive lanes for the entire 2.3 mile route
- signal pre-emption
- stations with large shelters and route information
- automatic vehicle location (AVL)
- next bus arrival information at kiosks
- new low-floor compressed natural gas (CNG) buses
- marketing and image development through vehicle graphics, stations, advertisements, and business tie-ins
- free fare, so no fare collection delay
The Lymmo route replaced an earlier downtown loop circulator, Freebee, which also charged no fare. One motivation behind the project was to provide a link from the Orange County Courthouse (opened 1997) and the 900 space Administration Parking Garage about half a mile away. The garage is the terminus of Lymmo, where buses lay over. The target market is people who drive to downtown Orlando and then use Lymmo to get to other locations, such as the Courthouse, restaurants, shopping, etc. The lack of connection to the rest of the Lynx system is exemplified by the fact that Lymmo does not stop at the downtown transit station, the destination of most routes in the system.
Lymmo was a substitute for the proposed use of historic trolleys to replace Freebee. The City of Orlando nixed the trolley idea because it would have had to contribute $20 million of its own funds. The total capital cost for Lymmo, $21 million, was half or less of what the trolley proposal would have cost.
Figure 1 Lymmo Right-of-Way and Station at Turn-Around Area
The exclusive lanes are paved with distinctive pavers. They are separated from general traffic lanes either with a raised median or a double row of raised reflective ceramic pavement markers embedded in the asphalt (see Figure 1). The route has loop sections at each end. In the middle segment, the two directions of Lymmo service are on the same street, with one of them being contraflow to the traffic lanes. For example, before reconstruction one portion of the route was a three-lane one-way street northbound. After reconstruction the street consists of, from right to left, on-street parking, northbound travel lane, raised median with bus stations, northbound bus lane, southbound contraflow bus lane, and sidewalk with bus stations. There was little or no opposition to establishment of bus lanes, perhaps because the affected streets had excess capacity, or adjacent streets could handle diverted traffic. On-street parking was eliminated on one side of some streets which form the Lymmo route.
Figure 2 Special Lymmo Signal Adjacent to Standard Signal
Because Lymmo operates in places and directions contrary to other traffic, all bus movements at intersections are controlled by special bus signals. To prevent confusion, these signal heads use lines instead of the standard red, yellow, and green lights (see Figure 1-2). When a bus approaches an intersection, a loop detector in the bus lane triggers the intersection to allow the bus to proceed either in its own signal phase (e.g. when making turns not otherwise permitted) or at the same time as other traffic is released when no conflicting traffic movements are permitted.
In the year following opening, ridership on Lymmo averaged about twice that of Freebee (see Figure 1-1 and Table 1-1). However, revenue hours of service supplied were 74% higher for Lymmo compared to Freebee, because service was extended into evenings and weekends.  The peak frequency of service, every five minutes, was not changed. The 2.3 mile Lymmo route is 25% shorter than the Freebee route (this is the round-trip distance, since it is a loop).
Figure 3 Lymmo Ridership Trend
Despite exclusive lanes and signal pre-emption, scheduled trip speeds are one-third lower on Lymmo than its predecessor. One explanation is that Lymmo buses stop at each station, whether a passenger has signaled a stop or not. Another possibility is that the increased ridership has resulted in more dwell time while passengers are boarding, despite the low floor vehicles and the absence of fare collection time. In fact, Table 1 also shows that average boardings per trip have increased from 15 to 20, 33%, on a 25% shorter route.
Figure 4 Lymmo Station with Next Bus LED Display
Waiting time might be lower for Lymmo compared to its predecessor, if dispatchers were to use the AVL system to instruct drivers to hold early buses to adjust to the schedule in order to achieve a more even distribution of headways. Because bunching may not have been much of a problem with Freebee (given that it was a short route with average loads well below capacity), there may not have been much bunching of buses before Lymmo service started. The new information kiosks provide information about next bus arrival, reducing the anxiety associated with waiting (see Figure 1-4).
Operating costs are primarily a function of total hours of service. Based on this measure, it is estimated that in 1998 Lymmo cost $1.2 million to operate, 65% more than the Freebee service of 1997. Because of the even greater increase in ridership, the average cost per boarding decreased from $1.37 for Freebee to $1.14 for Lymmo.
The total capital cost of the system was $21 million, of which $3 million was for vehicles, $0.4 million for landscaping, and the remainder for street reconstruction, shelters, information kiosks, AVL, traffic signals, banners, and other expenses. Assuming a discount rate of 7% and a lifetime of 40 years, the annualized capital costs are $1.6 million, or $1.43 per trip.
Figure 5 Lymmo Bus Painted in Leonardo DaVinci Theme
Orlando's Downtown Development Board worked with Lynx to market the project. Marketing pieces were sent to residents and merchants. When the Lymmo service started, the project was featured in a cover story in Downtown Orlando Monthly. Bus stations say "Lymmo" in large letters. The 10 new buses are painted with artistic themes, in what Lynx calls a moving museum or "Moveum" (see Figure 1-3). Every 6 to 12 months, Lynx expects to repaint its Lymmo bus fleet; it is seeking sponsors to pay the $100,000 cost. The initial series was based on "Imperial Tombs of China." The paint schemes are complemented by brochures inside the vehicles and banners on the road. Lymmo offers "Daily Deal" discounts from downtown merchants for riders who have a coupon stamped on board, which also enable riders to enter a raffle. Downtown Orlando events (e.g., festivals and concerts) have tie-ins to Lymmo service.
Much of the increased ridership of Lymmo compared to Freebee is due to the increased hours of service. However, the implied service elasticity of 1.35 (a 74% increase in revenue hours led to a 100% increase in ridership) is much larger than typical elasticities (often in the range of 0.3 to 0.7).  Another measure of the success of the project in attracting riders is the fact that average boardings per trip increased by one third.
Why did ridership increase? Despite the use of exclusive lanes and signal preemption, vehicle speeds decreased. Average waiting time might have been reduced due to the control of headways made possible by the AVL system, but this seems unlikely. The other possible sources for increased ridership (other than increased service):
- opening of the new courthouse
- better routing
- more pleasant service (stations and shelters, passenger information)
- aggressive marketing campaign
The Lymmo experience is not well suited to evaluating the potential impacts of exclusive lanes and signal preemption because there was no reduction in delay. In addition, the simultaneous implementation of a new route, new vehicles, greater hours of service, bus lanes, signal pre-emption, and marketing make it difficult to evaluate the independent effect of each change.
 The first two months of Lymmo service were not counted in the average because they were judged to be start-up months.
 Freebee operated 6 am to 7 pm weekdays. Lymmo extended service until 10 pm Monday to Thursday and until midnight Friday, and added service from 10 am to midnight Saturday and from 10 am to 10 pm Sunday.
 Travel speed was measured as revenue miles (excluding deadhead miles) divided by in-service hours (excluding recovery and deadhead time).
Mayworm et al. 1980 report the following average ridership elasticities with respect to vehicle miles of service from several studies in the 1970s: peak, 0.33, off-peak, 0.63, all hours, 0.69. Using a model of the U.S. transit industry as a whole, Schimek 1997 finds a service elasticity of 0.3 rising to 0.5 to 0.6 in 1.6 years.